Today's telecommunication technology continues to migrate the implementation of everyday applications and services onto computing networks, such as Internet Protocol (IP) networks. For instance, more and more users rely on multimedia conferencing that transports media data over computer networks using a variety of conferencing methods. A few examples of different types of conferencing methods include, but are not limited to Advance-Video Compression (AVC) multimedia conferencing, media relay conferencing (MRC), and mesh conferencing. In both AVC multimedia conferencing and MRC, a central entity (e.g., a traditional multipoint control unit (TMCU) or a media relay multipoint control unit (MRM), respectively) may manage the communication of control signals, audio media streams, and video media streams between one or more network devices. Each of the endpoints of a multimedia conference sends its media streams to the central entity that subsequently processes the media streams according to the type of communication method. The central entity then transfers the relevant media streams to the proper receiving endpoints. Alternatively, in mesh conferencing, no central entity may manage the different media streams. Instead, the mesh conferencing method may use a Roster List Server (RLS) as a central signaling entity and have the endpoints transmit and receive media streams directly from each other.
A multimedia conferencing session, such as a videoconferencing session, typically utilizes a sizeable amount of network resources to connect participants at different endpoints. For example, a multimedia conferencing session may need a relatively low latency and enough bandwidth to provide a real time experience for the participants. Further, a multimedia conferencing session may synchronize media streams (e.g., audio, video, and other media data) obtained from one or more endpoints. A multimedia conferencing session, for example, may match the different types of media and lip synchronization for audio and video streams obtained from participants at different endpoints. To synchronize media streams from multiple participants, a multimedia conferencing session may also attempt to limit the latency differences over the communication links. Additionally, if video data packets are lost in transmission, the decoded video may suffer from artifacts that may be repaired in order to maintain acceptable quality. Thus, missing data packets may lead to requesting Intra frames that can reduce a user experience (e.g., breathing effect) and increase bandwidth consumption.
In addition to allocating enough network resources, media conferencing methods generally involve a degree of network flexibility in order to adapt the media traffic to a variety of dynamic factors associated with a multimedia conferencing session. Typically, a multimedia conferencing session may be a dynamic session that varies traffic load over network links that carry the media streams. The dynamic nature of the multimedia conferencing session can originate from a variety of factors that include joining or leaving the session, changes in speakers, and changes in the presented images and mix audio. The different dynamic factors in the end often affect the routing of the different media streams and the addition or removal of network resources during the multimedia conferencing session. As such, improving media conferencing technology that is flexible in allocating network resources and capable of controlling and routing media streams according to the varying amounts of data traffic remains valuable as media conferencing becomes more popular.